| 神经网络在多传感器多目标跟踪中的应用 |
| |
| 引用本文： | 陈小惠,万德钧,王庆.神经网络在多传感器多目标跟踪中的应用[J].东南大学学报(自然科学版),2003,33(4):414-418. |
| |
| 作者姓名： | 陈小惠 万德钧 王庆 |
| |
| 作者单位： | 东南大学仪器科学与工程系,南京,210096 |
| |
| 基金项目： | 江苏省普通高校自然科学研究计划资助项目 ( 0 2KJB5 10 0 0 2 )，国防“十五”课题资助项目 ( 0 1F0 0 5 ) |
| |
| 摘 要： | 首先研究了基于粗关联和精关联过程的多传感器多目标(MSMT)跟踪融合算法，精关联是联合概率数据关联(JPDA)算法的推广，JPDA算法存在随传感器数和目标数的增加而计算量迅速增加的缺点；其次提出了一种基于神经网络的MSMT联合概率数据互联(MNJPDA)算法，MNJPDA算法能克服计算量爆炸问题，基于MNJPDA的融合算法能提高跟踪的快速性．仿真结果证明了MNJPDA融合算法的有效性．
|
| 关 键 词： | 多传感器多目标跟踪 数据融合 神经网络 数据关联 |
| 文章编号： | 1001-0505(2003)04-0414-05 |
Study on multisensor multitarget tracking using neural network |
| |
| Chen Xiaohui,Wan Dejun,Wang Qing.Study on multisensor multitarget tracking using neural network[J].Journal of Southeast University(Natural Science Edition),2003,33(4):414-418. |
| |
| Authors: | Chen Xiaohui Wan Dejun Wang Qing |
| |
| Abstract: | A fusion algorithm in multisensor multitarget (MSMT) tracking based on rough association and precise association is studied. The method of precise association is an extension of join probabilistic data association (JPDA) algorithm. The shortcoming of JPDA algorithm is that the computation load would increase evidently with the increase of sensor number and target number. A new multisensor multitarget JPDA algorithm based on neural network, named MNJPDA, is proposed. The MNJPDA algorithm can overcome the problem of expansion of computation load in MSMT tracking. The MNJPDA fusion algorithm can improve the tracking speed. The simulation results show the MNJPDA fusion algorithm is valid. |
| |
| Keywords: | multisensor multitarget tracking data fusion neural network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
